1. Field of the Invention
The present invention relates to a recording and/or reproducing apparatus, and more particularly to a recording and/or reproducing apparatus for digitally recording and/or reproducing information such as images and voices on/from a recording medium.
2. Description of the Related Art
Traditional apparatuses of this type include a digital video tape recorder (VTR) or digital video cassette recorder (VCR). In the digital VTR, there are two known digital recording methods: one is to record base band digital data associated with images or voices on a recording medium without data compression, and the other method records data, after reducing the amount of data, by performing data compression such as discrete cosine transformation (DCT) or variable length coding (VLC).
In the former method, or base band recording method, signals can be transmitted without degradation in the signal-to-noise ratio. In addition, in the base band recording method, it is possible to connect different types of apparatus to each other, and a digital switch is available. However, because the base band recording method does not involve performing data-compression, there are disadvantages in that a large amount of data must be recorded and that long recording on a recording medium is impossible.
In the latter method, on the other hand, signals (video or voice signals) to be recorded are subjected to high efficiency coding and further to error correction coding. The signals are then modulated and recorded on a recording medium. When the signals are reproduced, inverse processing is carried out. This method has the advantage that the amount of recording data can be greatly reduced. However, this method also has the disadvantage that distortion of data occurs due to the high efficiency coding and decoding, i.e., noise is generated and a slight degradation of picture quality occurs.
For example, when compression coding is carried out by means of orthogonal transformation such as DCT, data is rounded off after orthogonal transformation. Thus, the data digits cannot be perfectly preserved during calculation. These errors cause distortion of data after inverse orthogonal transformation. If editing or dubbing is repeatedly carried out, the degradation in picture quality will gradually increase.
To overcome this problem, a method has been proposed in which dubbing is carried out with coded signals before decoding. For example, in a reproducing apparatus, error correction is performed on reproducing data and if there is coded data having an error which cannot be corrected, interpolation is performed on the data using data corresponding to the same pixel location of preceding and succeeding frames. Then, the data is sent to an error correction coding circuit in the recording apparatus.
However, in the conventional recording and reproducing apparatus described above, the recording apparatus treats all data in the same way, without distinction between the data which has been interpolated in the reproducing apparatus and the data which has been reproduced correctly. The recording apparatus therefore records the data on a recording medium without distinguishing between the data which has been interpolated in the reproducing apparatus and the correct data which has not been interpolated.
As a result, when data which has been dubbed in this way is reproduced, the reproduction is performed without distinguishing between the data which has been interpolated and the data which has been reproduced correctly. If failure occurs in correction for some data, there is a possibility that the interpolation circuit will perform interpolation on this uncorrectable data by using the data which has already been interpolated by the reproducing apparatus during the dubbing operation. If such an interpolation occurs, the desired quality of pictures and sounds cannot be maintained and degradation will occur in the reproduced pictures.
In addition, the characteristic features of pictures do not contribute to interpolation for high-efficiency-coded data. This causes such a problem in that it is impossible to perform adaptive processing, such as switching between intra-frame interpolation (or intra-field interpolation) and inter-frame interpolation (or inter-field interpolation) depending on motion detection. As a result, in some cases, distortion occurs in a region corresponding to interpolated data.